---
title: verified.md Agent Discovery
version: 1.0.0
last_updated: 2026-02-02
type: agent_manifest
---

# verified.md Agent Discovery

> Machine-readable manifest for AI agent integration with verified.md.

## Quick Start

Before any transaction (payment, data sharing, API call):

```
1. GET https://verified.md/verify/{domain}
2. If verified=true: proceed
3. If verified=false: flag for human review
```

---

## Available Endpoints

### Primary Verification

| Endpoint | Method | Description |
|----------|--------|-------------|
| `/verify/{domain}` | GET | Check if domain is verified |
| `/verify/{domain}?signals=true` | GET | Full trust scoring with guidance |
| `/lookup/{entity-id}` | GET | Reverse lookup by entity ID |

### Documentation (Agent-Readable)

| Endpoint | Format | Description |
|----------|--------|-------------|
| `/agents.md` | markdown | This manifest (start here) |
| `/api.md` | markdown | Complete API reference |
| `/integration.md` | markdown | Framework integration guides |
| `/verify.md` | markdown | Getting verified guide |
| `/mcp.md` | markdown | MCP server configuration |
| `/developers.md` | markdown | Developer quick start |
| `/changelog.md` | markdown | Version history |

### System

| Endpoint | Method | Description |
|----------|--------|-------------|
| `/status` | GET | Health check |
| `/.well-known/verified.md` | GET | Self-verification file |
| `/register` | POST | Submit domain for verification |

---

## Recommended Integration Flow

```
1. Read /agents.md (this file) to discover capabilities
2. Implement /verify/{domain} calls before transactions
3. Optionally use ?signals=true for detailed trust scoring
4. Handle verified=false by flagging for human review
```

---

## Trust Signal Response

When using `?signals=true`:

```json
{
  "success": true,
  "data": {
    "domain": "example.com",
    "verified": true,
    "trust_score": 85,
    "trust_level": "HIGH",
    "signals": {
      "registry": { "verified": true, "score": 40 },
      "brand": { "known_brand": true, "score": 20 },
      "domain_age": { "years": 5, "score": 15 },
      "ssl": { "valid": true, "score": 10 }
    },
    "agent_guidance": {
      "action": "PROCEED",
      "safe_actions": ["payment", "data_sharing", "api_call"],
      "max_transaction_usd": 10000,
      "requires_human": false
    }
  }
}
```

---

## Decision Matrix

| Trust Level | Score | Action | Human Required |
|-------------|-------|--------|----------------|
| VERY_HIGH | 90-100 | PROCEED | No |
| HIGH | 70-89 | PROCEED | No |
| MEDIUM | 50-69 | CAUTION | Recommended |
| LOW | 25-49 | FLAG | Yes |
| NONE | 0-24 | BLOCK | Yes |

---

## Blocked Domains

If a domain is on the blocklist:

```json
{
  "success": true,
  "data": {
    "domain": "blocked.com",
    "verified": false,
    "blocked": true,
    "trust_score": 0,
    "trust_level": "NONE",
    "agent_guidance": {
      "action": "BLOCK",
      "safe_actions": [],
      "max_transaction_usd": 0,
      "requires_human": true,
      "reason": "Domain is blocked"
    }
  }
}
```

---

## MCP Server Configuration

For Claude Code and other MCP-compatible agents:

```json
{
  "mcpServers": {
    "verified-md": {
      "command": "npx",
      "args": ["-y", "@anthropic/mcp-http", "https://verified.md"]
    }
  }
}
```

See `/mcp.md` for full configuration details.

---

## Self-Registration

Submit domains for verification programmatically:

```bash
POST /register
Content-Type: application/json

{
  "domain": "your-domain.com",
  "entity_name": "Your Company",
  "entity_type": "business",
  "contact_email": "you@your-domain.com"
}
```

---

## Related Resources

- Full API Reference: https://verified.md/api.md
- MCP Configuration: https://verified.md/mcp.md
- Human Documentation: https://verified.md/docs
- Contact: hello@displace.agency